Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
QI: Fachverband Quanteninformation
QI 10: Certification and Benchmarking of Quantum Systems
QI 10.1: Vortrag
Donnerstag, 23. September 2021, 10:45–11:00, H5
Machine-learning framework for customized optimal quantum state tomography — •Violeta Ivanova-Rohling1,2,3, Guido Burkard1, and Niklas Rohling1 — 1Department of Physics, University of Konstanz, Germany — 2Zukunftskolleg, University of Konstanz — 3Department of Mathematical Foundations of Computer Sciences, Institute of Mathematics and Informatics, Bulgarian Academy of Sciences, Sofia, Bulgaria
Fastest quantum state tomography (QST) schemes which reach a desired precision are of high practical relevance. Rarely, analytical solutions are known, e.g. non-degenerate projective measurements whose eigenbases form a complete set of mutually unbiased bases (MUBs) [1]; mutually unbiased subspaces (MUSs) constructed from a complete set of MUBs if measuring one out of N qubits [2]. Our flexible scheme finds numerically an optimized QST measurement set given the system's specifications, e.g. for a qubit-qutrit system (e.g. NV center in diamond), a QST measurement set closely approximating MUSs [3]. Furthermore, machine learning approaches now for individual rank-1 measurements in eight dimensions [4] outperform standard numerical methods yielding high-performing measurement sets with complex structure and symmetries. Funded by Zukunftskolleg (U. Konstanz) and Bulgarian National Science Fund, contract No KP-06-PM 32/8
[1] Wootters, Fields, Ann. Phys. 191, 363 (1989).
[2] Bodmann, Haas, Proc. Amer. Math. Soc. 146, 2601 (2018).
[3] Ivanova-Rohling, Burkard, Rohling, arXiv:2012.14494.
[4] Ivanova-Rohling, Rohling, Cyb. Inf. Technol. 20 (6), 61 (2020).